I completely agree and I have tried this with my past Auralic Aries as well as my new MSB DAC Renderer module. Other streaming apps/methods simply sound better.
Bit perfect is just one thing you need for good sound. Low jitter is another and that’s what people are NOT talking about here. It’s my hypothesis that Roon, likely RAAT itself, is introducing jitter or something like it to cause the flattening of soundstage and loss of depth. I have heard it in my own system as well as others, and at dealers, using custom PCs as well as high end servers, Innuous as well as the Roon Nucleus Plus.
I can’t believe more people haven’t been able to hear this issue.
The thread title is about bypassing RAAT when RAAT is the pipeline. You can’t bypass it you need to determin what’s wrong with it, if anything and then correct it.
Whatever is causing what you are hearing, it isn’t jitter.
Wikipedia: jitter is the deviation from true periodicity of a presumably periodic signal.
RAAT is not presumably periodic, because the network (IP) that it rides on is not. The network is asynchronous, does not even guarantee that packets are delivered in the right order (they are sorted in software on arrival). There is no claim to exact timing is the software interface to the network stack.
I have given up on trying to identify the possible causes of audible differences between the various means of feeding output to my amplifier.
Roon was hosted on an old core i3 laptop (4010U) with only Roon as active application apart from windows 10 pro and its housekeeping processes.
This laptop died on me so I bought an intel NUC with an 8th gen core i5 processor and installed ROCK.
Sound quality has improved dramatically for some reason. Technically it doesn’t make any sense, but still it does sound different. My wife immediately commented on the improvement from the moment ROCK went online.
My suspicion is that, in cases where a change is made from WIN10 to a Linux based system, the perceived change in SQ is due to ASIO/WASAPI vs. ALSA drivers.
Not ROON/ROCK per se, but I guess the difference would be moot.
But wouldn’t those drivers only matter if using the internal sound card to produce the analog output. If the server is simply delivering the digital data to a separate endpoint, It seems these drivers would not be in the delivery chain.
I use multiple endpoints connected via network and those drivers always show up in Settings->Audio. If the drivers were only needed when using system output, then my DACs would register without installing any WIN10 vendor supplied drivers, which is not the case.
Ha, I’m not an expert by any means. Maybe some expert will chime in.
Since the creators of ROCK, and for that matter the Nucleus devices, have specifically stated that no SQ difference should be experienced by using these alternatives, and if one wanted to find an explanation other than expectation bias, then I am proposing that it’s because of different drivers.
During Xmas I had some more time to listen and experiment. I borrowed simple AC filter with 5 plugs and also some high-end power cables. And wow, the unpleasant tin sound of highs is gone. I still hear difference between streaming via ROON vs. directly to BluOS card, but now it is favorable to ROON, as it shows more details and sound is more “live”. The only drawback is that only well recorded tracks sound good. Many other recording suffer from unpleasant sibilance, which now my system presents ruthlessly. But I double checked this on headphones and this is just bad recording or mastering.
So, my apologies to ROON RAAT, the problem was somewhere else.
